Updates & fixes for Auto Leveler plugin #2022

nickmayer · 2022-10-29T17:33:08Z

AutoLeveler now warns about possible bugs on applying to gcode instead of spamming on startup
AutoLeveler apply to gcode will now remove old processors and replace instead of disabling button making it possible to use a different heightmap
Fix for MeshLeveler relative mode (G91)

… only)

codecov-commenter · 2022-10-29T17:40:48Z

Codecov Report

Merging #2022 (bd57fd7) into master (02990c0) will increase coverage by 0.02%.
The diff coverage is 40.00%.

@@             Coverage Diff              @@
##             master    #2022      +/-   ##
============================================
+ Coverage     22.51%   22.54%   +0.02%     
  Complexity     1997     1997              
============================================
  Files           577      577              
  Lines         24244    24252       +8     
  Branches       1922     1924       +2     
============================================
+ Hits           5459     5468       +9     
+ Misses        18314    18308       -6     
- Partials        471      476       +5

Impacted Files	Coverage Δ
...rsalgcodesender/gcode/processors/LineSplitter.java	`88.37% <0.00%> (-4.32%)`	⬇️
...lwinder/universalgcodesender/model/GUIBackend.java	`52.03% <33.33%> (ø)`
...ersalgcodesender/gcode/processors/MeshLeveler.java	`88.04% <45.45%> (-4.10%)`	⬇️
...ersalgcodesender/gcode/GcodePreprocessorUtils.java	`85.44% <100.00%> (+2.21%)`	⬆️
...llwinder/universalgcodesender/gcode/util/Code.java	`98.54% <0.00%> (+0.03%)`	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

winder · 2022-11-04T13:27:45Z

ugs-core/src/com/willwinder/universalgcodesender/gcode/processors/MeshLeveler.java

+        // If we're in relative mode there is no adjustment needed
+        // TODO: This won't be true if we start in relative mode, in which case we need to ensure the first command
+        // is to move down by the offset at (0, 0)
+        if (!state.inAbsoluteMode) {
+            return Collections.singletonList(commandString);


Is this really true? If there is a difference in the mesh heights between the current point and the previous one doesn't it need to be calculated and applied?

in other words, the current zPointOffset minus the previous zPointOffset

You're right. The previous code was wrong, but so is this. I'll send out an updated patch.

Updated. I was previously testing with gcode that only had relative z axis-only moves. This time I tested with the gcode below. It does still have an issue when the first move is relative, but I think it requires a broader change to be able properly detect and deal with this situation.

G91 G1 Z-1F300 G1 X10Z2F1000 G1 Y10Z2 G1 X-10Z-2 G1 Y-10Z-2

winder · 2022-11-04T13:35:08Z

...canner/src/main/java/com/willwinder/ugs/platform/surfacescanner/AutoLevelerTopComponent.java

@@ -599,26 +599,32 @@ private void dataViewerActionPerformed(java.awt.event.ActionEvent evt) {//GEN-FI
    }//GEN-LAST:event_dataViewerActionPerformed

    private void applyToGcodeActionPerformed(java.awt.event.ActionEvent evt) {//GEN-FIRST:event_applyToGcodeActionPerformed
-        GcodeParser gcp = new GcodeParser();


Was something broken with this approach? Without replacing the gcode parser you could have other processors loaded which might effect the final generated program.

I noticed that applyGcodeParser was deprecated and I assumed that having all processors that loaded run would be the expected behavior...wouldn't one expect those other processors to effect the final generated program?

I remember deprecating this method as I thought it got unintuitive when you applied a processor the previous one got replaced. For instance this prevented the user to apply "Mirror" and "Run from" at the same time.

It is still a bit unintuitive as the changes aren't reflected in the GCode which might have made more sense.

+1 to reflecting the changes in the GCode in some way. I considered adding a (read-only) view of what will actually be sent out when I was debugging this.

The default processors would remove comments and I always tried to maintain those in the editor view, but I can see how it's misleading to display different gcode than what gets sent.

Would it not be best to apply all processors in a serial fashion, rather than just the last? The order could be pre-determined based on what made the most sense.

…start AutoLeveler apply to gcode will now remove old processors and replace instead of disabling button Fixes for MeshLeveler relative mode (G91)

daxliniere · 2022-11-13T22:34:56Z

I would be happy to test this new auto-level function if you would like. Just let me know what you need me to do. 👍

nickmayer · 2022-11-14T00:50:04Z

FWIW, I've engraved several PCBs using this code with a 0.05mm depth and it has worked great for me. I do often forget to hit the apply button first though, so I may look into adding a reminder if you have scan data loaded but haven't applied it to your gcode 😉

daxliniere · 2022-11-14T23:08:35Z

I hate to be 'that guy', but are there directions somewhere on how to build this version into something I can run on my Windows machine, please?

daxliniere

Wow, fantastic. Nice work Nick and Joacim.
With the recent Github build script addition, does this mean that downloading the Nightly will include this patch?

EDIT: I downloaded the latest nightly (titlebar says 2022-12-28), but I still get the 'Autoleveler doesn't work' warning on startup. @nickmayer indicated here that it should only warn when autoleveling is applied to gcode, so I'm left wondering if this nightly is actually up to date.

breiler · 2022-12-29T06:03:52Z

@daxliniere I haven't merged it yet as I have not been able to test it (my workshop is temporarily used for storage due to some renovations so there is very limited access). You are more than welcome to test it and report any findings.

It should be available in the nightly build as soon as this finishes:
https://github.com/winder/Universal-G-Code-Sender/actions/runs/3798673377

daxliniere · 2022-12-29T22:58:39Z

Will do Joacim, but won't be back at the workshop for at least another week, I'd say.

daxliniere · 2022-12-29T23:03:59Z

Just extracted it and there's a problem. I got the usual issue that requires cache to be deleted. I deleted it in both locations and on re-loading I got 'module needed, but not found' errors. I clicked Exit, then on re-load again, it loads without error message, but looks like this:

There are no cache folders left.

daxliniere · 2022-12-30T01:55:25Z

Uhoh.. looks like the nightly borked my settings. :(
This is how my 2.0.13 looks now:

breiler · 2022-12-30T05:34:31Z

@daxliniere this is so weird that this happens to you. I upgrade/downgrade UGS all the time without this ever happening to me.

Anyway, since version 2.0.13 the cache folder is now located under:
<your user directory/AppData/Roaming/ugsplatform/var/cache

Sorry for the inconvenience... =(

daxliniere · 2022-12-30T05:43:15Z

Yup, I have deleted both cache dirs, but I still have this problem. Any idea if I can get my layout back?

daxliniere · 2022-12-30T05:44:36Z

I upgrade/downgrade UGS all the time without this ever happening to me.

How would I perform a clean installation? Maybe that will prevent this happening?

breiler · 2022-12-30T06:01:54Z

Any idea if I can get my layout back?

When the error occurred it hopefully just broke the window locations. But I am afraid that those settings are lost. As I have never been able to reproduce the error it would have been interesting to see your log file when this happens.

You should be able to restore the window positions by using the menu Windows / Reset Windows. It is probably better to try that first, otherwise you will risk losing your joystick mappings if they aren't already gone...

How would I perform a clean installation? Maybe that will prevent this happening?

A clean install would be to download and unzip UGS:

Make sure UGS is not running
Remove all files under C:/Users/[your username]/AppData/Roaming/ugsplatform
Remove all files under C:/Users/[your username]/.ugs (this is generally never needed)
Start ugs

lastacorn added 2 commits October 28, 2022 16:22

Fix bug with relative mode in LineSplitter

dfa6928

Update LineSplitter to pass through lines with no position data (feed…

c55a40f

… only)

winder reviewed Nov 4, 2022

View reviewed changes

Update AutoLeveler to only warn when applying to gcode instead of on …

bd57fd7

…start AutoLeveler apply to gcode will now remove old processors and replace instead of disabling button Fixes for MeshLeveler relative mode (G91)

nickmayer force-pushed the auto-leveler branch from eb10afa to bd57fd7 Compare November 4, 2022 15:43

daxliniere mentioned this pull request Nov 28, 2022

Add FluidNC probe support #2058

Merged

Merge branch 'master' into auto-leveler

692128c

daxliniere reviewed Dec 29, 2022

View reviewed changes

Merge branch 'master' into auto-leveler

99de391

breiler merged commit 497be97 into winder:master Dec 29, 2022

nickmayer deleted the auto-leveler branch February 18, 2023 17:45

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Updates & fixes for Auto Leveler plugin #2022

Updates & fixes for Auto Leveler plugin #2022

nickmayer commented Oct 29, 2022

codecov-commenter commented Oct 29, 2022 •

edited

Loading

winder Nov 4, 2022

winder Nov 4, 2022

nickmayer Nov 4, 2022

nickmayer Nov 4, 2022

winder Nov 4, 2022

nickmayer Nov 4, 2022

breiler Nov 4, 2022

nickmayer Nov 4, 2022

winder Nov 5, 2022

daxliniere Nov 14, 2022

daxliniere commented Nov 13, 2022

nickmayer commented Nov 14, 2022

daxliniere commented Nov 14, 2022

daxliniere left a comment •

edited

Loading

breiler commented Dec 29, 2022 •

edited

Loading

daxliniere commented Dec 29, 2022

daxliniere commented Dec 29, 2022

daxliniere commented Dec 30, 2022

breiler commented Dec 30, 2022

daxliniere commented Dec 30, 2022

daxliniere commented Dec 30, 2022

breiler commented Dec 30, 2022

Updates & fixes for Auto Leveler plugin #2022

Updates & fixes for Auto Leveler plugin #2022

Conversation

nickmayer commented Oct 29, 2022

codecov-commenter commented Oct 29, 2022 • edited Loading

Codecov Report

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

daxliniere commented Nov 13, 2022

nickmayer commented Nov 14, 2022

daxliniere commented Nov 14, 2022

daxliniere left a comment • edited Loading

Choose a reason for hiding this comment

breiler commented Dec 29, 2022 • edited Loading

daxliniere commented Dec 29, 2022

daxliniere commented Dec 29, 2022

daxliniere commented Dec 30, 2022

breiler commented Dec 30, 2022

daxliniere commented Dec 30, 2022

daxliniere commented Dec 30, 2022

breiler commented Dec 30, 2022

codecov-commenter commented Oct 29, 2022 •

edited

Loading

daxliniere left a comment •

edited

Loading

breiler commented Dec 29, 2022 •

edited

Loading